My dissertation research focused on interactions between birds and ornithophilic mosquitoes, both in the field and in a laboratory setting, by investigating aspects of mosquito feeding patterns and avian anti-mosquito defensive behavior.My first project was a two-year mosquito trapping study to evaluate the effects of tree canopy placement and bait bird species on capture rates of mosquitoes in upstate New York. I tested the hypothesis that mosquito vectors in the northeast prefer to feed on passerine birds in the tree canopy rather than chickens at ground-level, by comparing chicken- and sparrow-baited trap mosquito collections in the tree canopy (9m) and ground-level (1.5m). My results indicated that the sum of all ornithophilic mosquitoes species collected were captured more often in canopy traps compared with ground-level ones. Consequently, arbovirus surveillance could be improved by sampling both at ground-level and in the tree canopy.Mosquitoes showed no preference between bird species.
This was the first field study to separate Culex restuans from Cx. p. pipiens definitively with molecular markers. Cx. restuans was captured more often in the tree canopy, while Cx. p. pipiens was captured at similar rates at ground- and canopy-level. I also investigated escape rates of mosquitoes in the traps. I found that escape rates are highly variable and can be as high as 99%, a result that should be taken into consideration when evaluating lard-can trap data.
Based on preliminary field data indicating that mosquitoes feeding more successfully on sparrows than chickens, I tested the hypothesis that mosquitoes feed more successfully on house sparrows than chicks because house sparrows defend themselves less vigorously against mosquito attack. I characterized and quantified the most common anti-mosquito defensive behaviors in birds. Chicken defensive intensity appears to decline with age for younger chicks. In house sparrows, behavior intensity is proportional to mosquito density. Furthermore, even though there was considerable variance in the amount of sparrow defensive behavior, the blood feeding success of mosquitoes was uniformly high (80-85%), suggesting that behavior alone may not be effective in protecting individual birds from mosquito attack.
My last dissertation project was an investigation of the effect of disease on avian anti-mosquito defensive behavior and mosquito blood-feeding success. For these studies a pair of house finches was exposed to mosquito attack. Bird microsatellite DNA in the blood meal of each engorged mosquito was analyzed to determine host identity. One house finch in each pair was then infected with Mycoplasma gallisepticum, a poultry pathogen. Infected birds displayed less intense anti-mosquito behavior than healthy birds, but infected finches were actually fed upon less frequently than were healthy birds.
Interaction between experiment phase (x-axis) and status of house finch (control or treatment) as a predictor of total defensive behaviors. Treatment birds were infected during the infection phase: all birds were healthy at all other times. Behavior counts are given as least squares means in order to evaluate behavior after controlling for mosquito density, replicate number and cage position. Behaviors were square root-transformed to improve homoskedacity. P-values were adjusted using the Bonferroni correction.
